Showing papers on "Monte Carlo molecular modeling published in 1991"
01 Jan 1991
TL;DR: Markov chain Monte Carlo (MCMC) as discussed by the authors is a general tool for simulation of complex stochastic processes useful in many types of statistical inference, including maximum likelihood estimation and maximum pseudo likelihood estimation.
Abstract: Markov chain Monte Carlo (e. g., the Metropolis algorithm and Gibbs sampler) is a general tool for simulation of complex stochastic processes useful in many types of statistical inference. The basics of Markov chain Monte Carlo are reviewed, including choice of algorithms and variance estimation, and some new methods are introduced. The use of Markov chain Monte Carlo for maximum likelihood estimation is explained, and its performance is compared with maximum pseudo likelihood estimation.
1,242 citations
TL;DR: In this article, a class of multicanonical Monte Carlo algorithms is presented which can reduce the slowing down to a quadratic power law ≈V2. But this algorithm is not suitable for the case of finite volumes.
1,086 citations
TL;DR: If a ‘‘dynamical hierarchy’’ of transition probabilities is created which also satisfy the detailed‐balance criterion, then Monte Carlo methods may be utilized to simulate the Poisson process and both static and dynamic properties of model Hamiltonian systems may be obtained and interpreted consistently.
Abstract: Monte Carlo methods are utilized as computational tools in many areas of chemical physics. In this paper, we present the theoretical basis for a dynamical Monte Carlo method in terms of the theory of Poisson processes. We show that if: (1) a ‘‘dynamical hierarchy’’ of transition probabilities is created which also satisfy the detailed‐balance criterion; (2) time increments upon successful events are calculated appropriately; and (3) the effective independence of various events comprising the system can be achieved, then Monte Carlo methods may be utilized to simulate the Poisson process and both static and dynamic properties of model Hamiltonian systems may be obtained and interpreted consistently.
1,039 citations
Book•
22 Mar 1991
TL;DR: In this article, the authors present a book which discusses the same topics in the three levels known from the literature and gives useful information for both beginners and experienced readers, both well-established old techniques and also newest findings.
Abstract: With this book we try to reach several more-or-less unattainable goals namely: To compromise in a single book all the most important achievements of Monte Carlo calculations for solving neutron and photon transport problems. To present a book which discusses the same topics in the three levels known from the literature and gives us useful information for both beginners and experienced readers. It lists both well-established old techniques and also newest findings.
470 citations
Book•
01 Aug 1991
TL;DR: In this article, the DAMOCLES Monte Carlo Device Simulation Program (DMCDPS) is implemented for Semiconductor Heterostructure Devices and Monte Carlo Simulation of Quasi-One-Dimensional Systems.
Abstract: 1. Numerical Aspects and Implementation of the DAMOCLES Monte Carlo Device Simulation Program.- 2. Scattering Mechanisms for Semiconductor Transport Calculations.- 3. Evaluating Photoexcitation Experiments Using Monte Carlo Simulations.- 4. Extensions of the Monte Carlo Simulation in Semiconductors to Fast Processes.- 5. Theory and Calculation of the Deformation Potential Electron-Phonon Scattering Rates in Semiconductors.- 6. Ensemble Monte Carlo Investigation of Nonlinear Transport Effects in Semiconductor Heterostructure Devices.- 7. Monte Carlo Simulation of Quasi-One-Dimensional Systems.- 8. The Application of Monte Carlo Techniques in Advanced Hydrodynamic Transport Models.- 9. Vectorization of Monte Carlo Algorithms for Semiconductor Simulation.- 10. Full Band Monte Carlo Program for Electrons in Silicon.
285 citations
CERN1
TL;DR: A hamiltonian-guided Monte Carlo algorithm for simulations of lattice field theories allowing the trajectory length to be shrunk to the step-size without losing on the speed of configuration decorrelation is presented.
279 citations
Book•
01 Oct 1991
TL;DR: General Schemes for Constructing Scalar and Vector Monte Carlo Algorithms for Solving Boundary Value Problems: Random Walks on Boundary and Inside the Domain Algorithm and Numerical Experiments.
Abstract: General Schemes for Constructing Scalar and Vector Monte Carlo Algorithms for Solving Boundary Value Problems: Random Walks on Boundary and Inside the Domain Algorithms. Random Walks and Approximations of Random Processes. Monte Carlo Algorithms for Solving Integral Equations: Algorithms Based on Numerical Analytical Continuation. Asymptotically Unbiased Estimates Based on Singular Approximation of the Kernel. The Eigen-Value Problems for Integral Operators. Alternative Constructions of the Resolvent: Modifications and Numerical Experiments. Monte Carlo Algorithms for Solving Boundary Value Problems of the Potential Theory: The Walk on Boundary Algorithms for Solving Interior and Exterior Boundary Value Problems. Walk Inside the Domain Algorithms. Numerical Solution of Test and Applied Problems of Potential Theory in Deterministic. Monte Carlo Algorithms for Solving High-order Equations and Problems in Elasticity: Biharmonic Problem. Metaharmonic Equations. Spatial Problems of Elasticity Theory. Applications to Stochastic Elasticity Problems. Diffusion Problems: Walk on Boundary Algorithms for the Heat Equation. The Walk Inside the Domain Algorithms. Particle Diffusion in Random Velocity Fields. Applications to Diffusion Problems.
263 citations
TL;DR: The details of an application of the method of maximum entropy to the extraction of spectral and transport properties from the imaginary-time correlation functions generated from quantum Monte Carlo simulations of the nondegenerate, symmetric, single-impurity Anderson model are reported.
Abstract: We report the details of an application of the method of maximum entropy to the extraction of spectral and transport properties from the imaginary-time correlation functions generated from quantum Monte Carlo simulations of the nondegenerate, symmetric, single-impurity Anderson model. We find that these physical properties are approximately universal functions of temperature and frequency when these parameters are scaled by the Kondo temperature. We also found that important details for successful extractions included the generation of statistically independent, Gaussian-distributed data, and a good choice of a default model to represent the state of our prior knowledge about the result in the absence of data. We suggest that our techniques are not restricted to the Hamiltonian and quantum Monte Carlo algorithm used here, but that maximum entropy and these techniques lay the general groundwork for the extraction of dynamical information from imaginary-time data generated by other quantum Monte Carlo simulations.
242 citations
TL;DR: In this paper, a number of ways of calculating exact Monte Carlo p-values by sequential sampling are discussed. But, in particular, a sequential method is proposed for dealing with situations in which values can only be conveniently generated using a Markov chain, conditioned to pass through the observed data.
Abstract: SUMMARY The assessment of statistical significance by Monte Carlo simulation may be costly in computer time. This paper looks at a number of ways of calculating exact Monte Carlo p-values by sequential sampling. Such p-values are shown to have properties similar to those obtained by sampling with a fixed sample size. Both standard and generalized Monte Carlo procedures are discussed and, in particular, a sequential method is proposed for dealing with situations in which values can only be conveniently generated using a Markov chain, conditioned to pass through the observed data.
213 citations
01 Jan 1991-International Journal of Radiation Applications and Instrumentation. Part A. Applied Radiation and Isotopes
TL;DR: The Monte Carlo code ETRAN as mentioned in this paper was developed for the solution of coupled electron-photon transport problems, and it has been used extensively in the field of wireless sensor networks.
171 citations
TL;DR: The core of the approach is introducing network time-evolution processes and using certain graph-theoretic machinery, resulting in a considerable increase in accuracy for Monte Carlo estimates, especially for highly reliable networks.
Abstract: Monte Carlo techniques for estimating various network reliability characteristics, including terminal connectivity, are developed by assuming that edges are subject to failures with arbitrary probabilities and nodes are absolutely reliable. The core of the approach is introducing network time-evolution processes and using certain graph-theoretic machinery, resulting in a considerable increase in accuracy for Monte Carlo estimates, especially for highly reliable networks. Simulation strategies and numerical results are presented and discussed. >
TL;DR: Results show that Monte Carlo filtering with a behavior definition that is closely related to the sensitivity structure of the model will produce substantial reductions in model forecasting uncertainty.
Abstract: Complex models are often used to make predictions of environmental effects over a broad range of temporal and spatial scales. The data necessary to adequately estimate the parameters of these complex models are often not available. Monte Carlo filtering, the process of rejecting sets of mode! simulations that fail to meet prespecified criteria of model performance, is a useful procedure for objectively establishing parameter values and improving confidence in model predictions. This paper uses a foodweb model to examine the relationship between model sensitivities and Monte Carlo filtering. Results show that Monte Carlo filtering with a behavior definition that is closely related to the sensitivity structure of the model will produce substantial reductions in model forecasting uncertainty.
TL;DR: In this article, it is shown that the Metropolis Monte Carlo scheme contains a description of the physical diffusion process and that the use of the technique is not to be restricted to its conventional application for studies of the equilibrium properties of fluids but should be extended to studies of their dynamic properties.
TL;DR: A review of the theory of chemical bonding based on replacement of an N-atom system by N individual systems each consisting of an atom embedded in a homogeneous electron gas can be found in this article.
Abstract: We review recent developments in the theory of chemical bonding based upon replacement of an N-atom system by N individual systems each consisting of an atom embedded in a homogeneous electron gas. These theories include the corrected effective medium and effective-medium-based methods, which are either first principle or semi-empirical, as well as the embedded atom and related methods (e.g. the “glue” and Finnis-Sinclair methods), which are totally empirical. These methods can provide an accurate description of metal-metal interactions for simple or transition metals with weak d bonding, including homogeneous and heterogeneous systems. They also can describe the binding of non-metallic atoms to metals. A number of these methods are efficient enough computationally to be used in molecular dynamics and/or Monte Carlo simulations of systems with many thousands of atoms.
TL;DR: In this article, a Monte Carlo coarse graining approach was proposed to deduce macroscopic properties of specific polymer melts using a Monte-Carlo coarse grained approach. But this approach was not suitable for the case of Bisphenol A.
Abstract: A new approach to deduce macroscopic properties of specific polymer melts using Monte Carlo coarse graining is proposed. Distribution functions for bond lengths and angles of chemically realistic single chains are used in input. As a first application the Vogel-Fulcher temperature for Bisphenol A polycarbonate is predicted.
TL;DR: It is shown that, in locating an optimal model, the new technique is far superior in performance to Monte Carlo techniques in all cases considered, and Monte Carlo integration is still regarded as an effective method for the subsequent model appraisal.
Abstract: In providing a method for solving non-linear optimization problems Monte Carlo techniques avoid the need for linearization but, in practice, are often prohibitive because of the large number of models that must be considered. A new class of methods known as Genetic Algorithms have recently been devised in the field of Artificial Intelligence. We outline the basic concept of genetic algorithms and discuss three examples. We show that, in locating an optimal model, the new technique is far superior in performance to Monte Carlo techniques in all cases considered. However, Monte Carlo integration is still regarded as an effective method for the subsequent model appraisal.
TL;DR: In this article, trajectories of the folding pathways of α-helical hairpin proteins have been computed by two very different models and simulation schemes, and an examination of the resulting pathways suggests that the on-site mechanism of assembly previously found in Monte Carlo diamond lattice simulations holds in general for the initial stages of protein folding.
TL;DR: A general purpose Monte Carlo computer code for the calculation of radiative exchange factors in three-dimensional enclosures with a nonparticipating medium is described in this article, where the effect of input file parameters on computer lime usage is investigated, and rates of convergence are calculated.
Abstract: A general purpose Monte Carlo computer code for the calculation of radiative exchange factors in three-dimensional enclosures with a nonparticipating medium is described. Capabilities of the code include mixed specular and diffuse reflection models, banded spectral material properties, transmission through external surfaces, and simulation of beam radiation. The ray tracing and surface interaction routines are described, as well as the shading algorithms. The effect of input file parameters on computer lime usage is investigated, and rates of convergence are calculated. Results for run time and accuracy are in good agreement with predictions.
TL;DR: In this paper, the predicted low-temperature, low-density phase separation of Coulombic systems has been sought by applying the recently developed Monte Carlo technique of "density scaling" to a fluid of charged spheres.
Abstract: The predicted low‐temperature, low‐density phase separation of Coulombic systems has been sought by applying the recently developed Monte Carlo technique of ‘‘density scaling’’ to a fluid of charged spheres. The existence of the transition is confirmed. As well as locating it, the method allows estimates of the thermodynamics of the transition.
TL;DR: In this article, the properties of rare gas atoms were calculated using variational Monte Carlo methods, including shadow, multiple bond length, and Boltzmann-like wave functions, and the results indicated delocalized helium and localized argon and neon.
Abstract: Clusters of rare gas atoms provide an interesting setting for the study of the issue of quantum mechanical localization. The properties of these clusters of 2–7 atoms are calculated using variational Monte Carlo methods. To our knowledge, this is the first variational Monte Carlo study of localized clusters and new solidlike wave function forms, including shadow, multiple bond length, and Boltzmann‐like wave functions, are reported. Diffusion Monte Carlo methods provide an independent, exact value of the ground state energy, useful as a check of the variational results. The properties of the variational wave functions, when analyzed in terms of probability distribution functions, quench studies, and visual examination of the wave functions, indicate delocalized helium and localized argon and neon clusters.
20 May 1991
TL;DR: In this article, the authors present the results of an analytic study of the Hybrid Monte Carlo algorithm for free field theory and calculate the acceptance rate and autocorrelation function as a function of lattice volume, integration step size, and (average) trajectory length.
Abstract: We present the results of an analytic study of the Hybrid Monte Carlo algorithm for free field theory We calculate the acceptance rate and autocorrelation function as a function of lattice volume, integration step size, and (average) trajectory length We show that the dynamical critical exponent z can be tuned to unity by a judicious choice of average trajectory length
TL;DR: In this paper, the authors demonstrate the feasibility of using "umbrella sampling" to do Monte Carlo Markov sampling runs each covering a substantial range of density: "density-scaling Monte Carlo", or DSMC.
TL;DR: In this paper, a method for carrying out Monte Carlo calculations for condensed-matter systems in the microcanonical ensemble is formulated and illustrated with example calculations, and a complete statistical mechanics associated with the Monte Carlo procedure is presented.
Abstract: A method for carrying out Monte Carlo calculations for condensed-matter systems in the microcanonical ensemble is formulated and illustrated with example calculations. A complete statistical mechanics associated with the Monte Carlo procedure is presented. The same method applies to any of the microcanonical-like shell ensembles.
TL;DR: In this article, a Monte Carlo convolution method for simulating time correlated single photon counting data is presented, which automatically produces the Poisson statistics of the real experiment, including pulse pileup.
Abstract: A Monte Carlo convolution method for simulating time‐correlated single photon counting data is presented. The random convolution automatically produces the Poisson statistics of the real experiment. The new simulation technique offers realistic treatment of various aspects of the single photon counting experiment, including pulse pileup. The random convolution is also incorporated in a data analysis technique using a reference fluorophore. Illustrative examples comparing the Monte Carlo and conventional simulation methods are given and the conceptual differences are discussed.
TL;DR: In this paper, the authors show that the choice of optimization functional can have a significant influence on the accuracy of variational Monte Carlo calculations and demonstrate that the Monte Carlo analog of the Rayleigh-Ritz procedure, which explicitly orthogonalizes ground and excited states, can be used to produce accurate eigenvalue and variance estimates of excited states.
Abstract: Using several simple systems as examples, we show that the choice of optimization functional can have a significant influence on the accuracy of variational Monte Carlo calculations. In addition, we demonstrate that the Monte Carlo analog of the Rayleigh–Ritz procedure, which explicitly orthogonalizes ground and excited states, can be used to produce accurate eigenvalue and variance estimates of excited states.
TL;DR: Partial radial distribution functions of standard Monte Carlo simulations were used as input for Reverse Monte Carlo simulation, a novel method for structural modeling as mentioned in this paper. But it has not been shown that sufficiently close agreement in radial distribution function involves deeper structural similarities.
Abstract: Partial radial distribution functions of standard Monte Carlo simulations were used as input for Reverse Monte Carlo simulation, a novel method for structural modeling. From detailed comparison of the two independent (MC and RMC) particle configurations it has turned out that sufficiently close agreement in radial distribution functions involves deeper structural similarities. RMC results, i.e., particle configurations produced by RMC upon the basis of partial radial distribution functions, are therefore applicable for detailed analysis of the structure, giving new prospects for data evaluation in neutron diffraction method. As a result of extensive Reverse Monte Carlo simulations of this work some new aspects of the technique itself could be revealed.
TL;DR: A new fully implicit Monte Carlo method for solving the nonlinear radiative transfer equations is presented, which is both temperature and frequency implicit, and the particle tracking is much faster than in the standard Fleck method.
Abstract: In this article, a new fully implicit Monte Carlo method for solving the nonlinear radiative transfer equations is presented. This scheme is both temperature and frequency implicit, and the particle tracking is much faster than in the standard Fleck method. This tracking is used to build an energy balance matrix. This leads to an original matrix formulation of the energy balance equation, where the entries of the matrix are computed by a Monte Carlo method.This matrix formulation of the energy balance equation allows this method to be easily coupled with the Rosseland diffusion equation, which is generally needed in optically thick media, via domain decomposition. This coupling scheme between the two methods is also fully implicit.Numerical tests are shown to prove the efficiency of the matrix Monte Carlo method; the results are better than those obtained by Fleck's method in that there are smaller computation times and there is a better robustness of the time discretization.